A key project in data administration is extraction and translation of data from one software and hardware system architecture to another. Usually the previous, or legacy, system programs and data are stored as third or fourth generation language applications, i.e. COBOL, which include some form of relational database, i.e. CICS. In the volatile environment in which businesses operate in today, these legacy systems are quickly converted to much higher level languages, i.e. Action Diagram Language, which includes data item represented as data models stored in an Information Repository. The newer systems are categorized as Computer Aided System Engineering (CASE) tools and integrate program generation and information storage with a user graphical friendly interface. The information repository, or encyclopedia, associated with a CASE tool integrates the storage of data including program files, graphs, diagrams and data structures. One such CASE tool is the Information Engineering Facility (IEF), a product of the assignee, Texas Instruments, Inc.
Toolsets provided by the IEF include Current System Analysis (CSA) and Current System Modification (CSM). The CSA is a tool for performing system wide analysis on COBOL systems and for performing inventory impact and "where used" and "how used" reporting on data items within the COBOL system. The CSM is a tool for modifying COBOL systems.
Currently the CSA tool uses an Encyclopedia Import Facility (ENCI) to capture legacy system information as data models for input to the Information Repository included in the IEF. These data models may include program code and diagrams. The only diagrams currently supported by the ENCI are entity relationship diagrams (ERD), activity hierarchy diagrams (AHD), dialog flow diagrams (DFD), and Screen Designs. All legacy systems, however, are not fully represented using one of the currently supported diagrams. Also, ENCI cannot be extended from batch oriented processing to on-line processing as required, for example, for interactively creating the data models from the legacy system code. Furthermore, the use of ENCI occurs early in the re-engineering project, during initial analysis of the legacy system. As the analysis continues, however, inconsistencies are found which require additional, unanticipated changes to the Information Repository which may affect previous translation results. Also, ENCI merely builds new data models to import to the Information Repository. If the data already exists or corresponds with data already in the Information Repository, then the ENCI generates data redundancy problems which must be addressed later.
Thus, what is needed is an automated, interactive translation tool to capture business rules and data items embedded in source, legacy, application system program logic, i.e. COBOL, cross reference the source system data items with data items on a target application system, then translate the extracted business rules and data items to an equivalent target application system representation in accordance with the cross reference for input to an Information Repository.